l6566bh based flyback converter ground short together

[Kaleeswaran]

hi all,

     i am designing flyback converter based on the l6566bh. i have 2 outputs on the secondary side, both are separate windings (different ground potential) in transformer. can i short together for shared ground reference if isolation is not necessary.

[Kaleeswaran]

Hi sir ,

Could you please review my schematic and point out any mistakes? I am observing ripple voltages of around 2 V to 2.5 (625khz frequency range)  on both the 15 V and 12 V outputs, but I am expecting ripple levels below 1 V."

 my specifications:

           input voltage 180-275vac, output power 81w

           output voltage_1 -> 15v, 3A rating( regulated by TL431)

           output voltage_2->12v, 3A rating,

           Reflected voltage -90v

          Switching freq -53khz,fixed frequency converter.

Transformer specifications:

           primary inductance-391uh  10% tolerance, Leakage inductance- 4.5uh

           number of primary turns- 34, Secondary 1 turns- 6, Secondary 2 turns-5  

          auxiliary turns – 7

          turns ratio- 5.66

Transformer specifications:

           primary inductance-391uh  10% tolerance,

           Leakage inductance- 4.5uh

           number of primary turns- 34

          Secondary 1 turns- 6

          Secondary 2 turns-5

          auxiliary turns – 7

          turns ratio- 5.66

[Peter BENSCH]

The schematics are hard to decipher with the current resolution of 1020x706, with double the resolution it would be a bit easier to look at it in depth.

As I mentioned before, too much ripple can have many causes, the component parameters are only one of them. The transformer design and especially the layout also play an important role. And as also mentioned, such ripple can be attenuated by passive LC filters in the output lines 12V and 15V.

[Kaleeswaran]

Hi Sir,

  Thank you for your reply. Please find the attached schematic in PDF format. I have also included an LC filter at the output, but it does not fully eliminate the noise. However, when I measure the 15V and 12V outputs using a ground spring instead of the oscilloscope’s alligator ground clip, the noise is significantly reduced, and the voltages appear much more stable. Why does this happen when using the ground spring? Is it appropriate to use a ground spring for accurate noise measurements?

[Peter BENSCH]

LC filters have a finite attenuation, so you can't completely avoid noise.

With the wire loop of the alligator ground clip you probably catch magnetic fields from the transformer or the output coils so that you see larger spikes. Your favourite search engine will find some discussions while searching for e.g. 'alligator ground clip vs spring'.

[Kaleeswaran]

Hi sir,

Thank you for your reply, Sir. Kindly review my schematic. If you would like to see the layout file as well, I am happy to share it.

[Peter BENSCH]

After reviewing the schematics, I don't notice any major problems, just:

• AC_OK and VFF can be derived from a common voltage divider (see data sheet, section 5.12, fig. 26), so you could combine R222...225 and R226...230 (not a mistake, just unnecessarily complex)

If you can measure better with the ground spring, an assessment of the layout is probably not necessary.

Regards
/Peter